Validation of Dosimetric and Non-Dosimetric Stability In Varian Eclipse TPS Following a Version 18 Upgrade: Establishing an Annual Quality Assurance Protocol
Abstract
Purpose
To evaluate the dosimetric and non-dosimetric stability of the Varian Eclipse Treatment Planning System (TPS) following a major software upgrade from v15 to v18. This study is to confirm that the system maintains the integrity of previously established v18 commissioning benchmarks, specifically following the introduction of the Enhanced Leaf Model (ELM), to ensure continued clinical suitability.
Methods
Stability assessments were performed for the Analytical Anisotropic Algorithm (AAA), Acuros XB (AXB), and Electron Monte Carlo (eMC) algorithms. Dosimetric stability was evaluated by comparing current calculations against established v18 commissioning benchmarks, including Reference Condition Plans (standard open field calculations vs. measurements, 1% tolerance) and Commissioned Test Plans (re-calculation of simple and complex plans vs. original results, 2% tolerance). Independent dose verification was conducted using Radformation’s ClearCalc that utilizes Finite-Size Pencil Beam algorithm (5% tolerance). Non-dosimetric testing evaluated the consistency of DICOM image transfer, geometric fidelity (2 mm tolerance), and the persistence of Hounsfield Unit (HU)-to-density table associations via DICOM header identification.
Results
All dosimetric tests fell within established tolerances. Maximum observed differences were 0.8% for reference condition plans, 0.1% for simple commissioned plans, and 1.5% for complex commissioned plans. Independent dose verification showed maximum deviations of 1.0% and 3.9% for reference and complex plans, respectively. Non-dosimetric tests demonstrated accurate image transfer, and correct HU-density table association. Geometric accuracy was maintained within a maximum discrepancy of 1.6 mm.
Conclusion
Eclipse TPS v18 demonstrated stable and reliable performance across all dosimetric and non-dosimetric QA metrics when compared to measurements and commissioned benchmarks. All calculated dose points were within tolerance for all algorithms and test plans. These verification procedures have been adopted as the institutional internal annual TPS quality assurance protocol.